基于机器视觉的电主轴轴承损伤识别与检测研究  

Research on Damage Identification and Detection of Electric Spindle Bearings Based on Machine Vision

作　　者：罗良传 郭伟科 张华伟 陈浩 王翔 廖菲 LUO Liangchuan;GUO Weike;ZHANG Huawei;CHEN Hao;WANG Xiang;LIAO Fei(Institute of Intelligent Manufacturing,Guangdong Academy of Sciences,Guangzhou Guangdong 510075,China;Guangdong Key Laboratory of Modern Control Technology,Guangzhou Guangdong 510075,China)

机构地区：[1]广东省科学院智能制造研究所,广东广州510075 [2]广东省现代控制技术重点实验室,广东广州510075

出　　处：《机床与液压》2024年第20期130-135,共6页Machine Tool & Hydraulics

基　　金：广东省科学院发展专项资金项目(2022GDASZH-2022010108);广州市重点领域研发计划“人工智能”重大科技专项项目(202206030006);广州市制造业绿色发展科技协同创新中心建设项目(2023B04J0022);2023年超精密高速气体静压电主轴项目(2023ZY01077);广州市科技计划项目重点研发计划(2023B04J0029)。

摘　　要：轴承是电主轴的核心回转部件,其稳定可靠的运行是确保电主轴正常运行的关键。针对高速电主轴核心回转部件轴承表面的裂纹损伤、润滑油污染等损伤识别检测难度大等问题,提出基于机器视觉的电主轴轴承损伤识别与检测系统。基于OpenCV库,对采集到的电主轴回转部件角接触球轴承的图像进行特征分类与形态学预处理以及损伤特征图像的灰度化处理,并对轴承损伤特征图像进行标记以及边缘特征提取,基于Visual Studio自主开发的轴承损伤特征识别算法对分类的轴承损伤特征进行识别和检测,实现了对电主轴轴承损伤的在线实时识别与检测,且具有较高的识别速度与检测率,有一定的实际应用价值。Bearings are the core components of the electric spindle,and their stable and reliable operation is the key to ensure the normal operation of the electric spindle.In response to the difficulties in identifying and detecting surface cracks and lubricating oil pollution on the bearings of high-speed electric spindle core rotating components,a machine vision based electric spindle bearing damage identification and detection system was proposed.Based on OpenCV library,the collected images of angular contact ball bearings of the electric spindle rotating components were preprocessed by feature classification and morphology and grayscale processing of damage feature images,the damage bearing feature images were marked,and edge features were extracted,based on Visual Studio self-developed bearing damage feature identification algorithm,the classified bearing damage features were identified and detected.The online real-time identification and detection of electric spindle bearing damage is realized,and it has high identification speed and detection rate,which has certain practical application value.

关 键 词：电主轴轴承 机器视觉 损伤识别与检测 

分 类 号：TG502.1[金属学及工艺—金属切削加工及机床]